Arm backend: fix(arm): validate partitions for dependency cycles after Q/DQ de-tagging

[beomwookang]

Fixes #18190

Summary

_detag_boundary_nodes() removes Q/DQ nodes from partition boundaries after CapabilityBasedPartitioner has produced cycle-free partitions.
However, this de-tagging can introduce dependency cycles for models with complex attention blocks (e.g. MobileViT, where CNN and Transformer ops are grouped into a single large partition).

The cycle occurs because removing Q/DQ bridge nodes creates paths that exit the partition and re-enter it through the now-unpartitioned nodes, making it impossible to extract the partition as a valid subgraph:

[partition] Linear_Q → [de-tagged Q] → [outside] → [de-tagged DQ] → [partition] Matmul
[partition] Linear_K → [de-tagged Q] → [outside] → [de-tagged DQ] → [partition] Matmul

Changes

After _detag_boundary_nodes(), validate each partition for dependency cycles.
When a cycle is detected, split the partition into connected components of the surviving tagged nodes.
Each component becomes a separate valid partition.

• Add _validate_partition(): BFS-based cycle detection (same algorithm as torch.fx.passes.utils.fuser_utils.validate_partition)
• Add _find_connected_components(): undirected graph traversal to split surviving nodes into disjoint sub-partitions
• Guard the nocompute-partition tags.remove() against already-removed tags from the cycle-split path

Test Results

• MobileViT-S on Ethos-U85: previously failed with AssertionError: Invalid partition, found dependency cycles, now successfully produces a .pte file (5.7 MB). Nine attention-block partitions are each split into 3 sub-partitions. All sub-partitions remain on NPU (no CPU fallback).
• CNN-only models (ResNet, MobileNetV2, EfficientNet): unaffected — no cycles after de-tagging, no partition splitting occurs.

How to test

import torch
from executorch.backends.arm.ethosu import EthosUPartitioner, EthosUCompileSpec
from executorch.exir import to_edge_transform_and_lower, EdgeCompileConfig

# Load a MobileViT (or any CNN+Transformer hybrid) model
exported_program = torch.export.load("mobilevit_s.pt2")
model = exported_program.module()
example_inputs = (torch.randn(1, 3, 256, 256),)

compile_spec = EthosUCompileSpec("ethos-u85-256")
partitioner = EthosUPartitioner(compile_spec)

# This previously raised: AssertionError: Invalid partition, found
dependency cycles
edge = to_edge_transform_and_lower(
    exported_program,
    partitioner=[partitioner],
    compile_config=EdgeCompileConfig(_check_ir_validity=False),
)

cc @digantdesai @freddan80 @per @zingo @oscarandersson8218 @mansnils @Sebastian-Larsson @robell

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18191

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 5 New Failures, 1 Cancelled Job

As of commit 5f597ab with merge base abc0237 ():

NEW FAILURES - The following jobs have failed:

• Apple / build-benchmark-app / macos-job (gh)
• pull / android / run-emulator (gh)
  The process '/usr/local/lib/android/sdk/platform-tools/adb' failed with exit code 224
• pull / test-phi-3-mini-runner-linux / linux-job (gh)
  RuntimeError: Command docker exec -t ec42ce8ef3e9d243cad75e13c4beb435b0b58325a235536ed402e16079650036 /exec failed with exit code 8
• trunk / test-multimodal-macos (gemma3-4b) / macos-job (gh)
  RuntimeError: Command bash /Users/runner/work/_temp/exec_script failed with exit code 1
• trunk / unittest-release / macos / macos-job (gh)
  export/tests/test_target_recipes.py::TestTargetRecipes::test_linear_model

CANCELLED JOB - The following job was cancelled. Please retry:

• trunk / test-mcu-cortex-m-backend / linux-job (gh)
  ##[error]The operation was canceled.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[github-actions]

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

[pytorch-bot]

To add the ciflow label ciflow/trunk please first approve the workflows that are awaiting approval (scroll to the bottom of this page).

This helps ensure we don't trigger CI on this PR until it is actually authorized to do so. Please ping one of the reviewers if you do not have access to approve and run workflows.